| Summary: | Global Positioning System (GPS) or currently upgraded to Global Navigation Satellite System (GNSS) has been applied in many SHM systems of the super-structures, especially in the long-span bridges. A GNSS system has the ability in monitoring the global deformation of a long-span cable-stayed bridge at the millimeter level of accuracy in real-time. However, the GNSS monitoring dataset acquired from a SHM system includes various noise data such as abnormal data, missing data, and so on. This paper studies de-noising methods to detect and replace the abnormal data of a GPS monitoring dataset acquired from a real cable-stayed bridge in Vietnam. Firstly, a GPS monitoring dataset of an actual long-span cable-stayed bridge was acquired to study processing abnormal data. A scenario of abnormal data was created in a time-series GPS data, and then the Hampel identifier method was applied to detect and replace the abnormal data. The replacing data were then assessed for precision and reliability by using correlation analysis and RMSE criterion. Finally, a long-term GNSS monitoring dataset processed the abnormal data automatically. The results show, that abnormal data in GPS monitoring data can be detected and replaced with high accuracy and reliability.
|
|---|